QUALITY OF SOIL SIMULATION BY INM RAS-MSU LAND SURFACE SCHEME AS A PART OF THE SL-AV WEATHER PREDICTION MODEL

Abstract

In situ soil moisture and temperature observations are used to evaluate soil moisture and temperature analysis of the global atmospheric model SL-AV for the summer season of 2014 across Western Europe. It is shown that the seasonal temperature analysis trends are reproduced better than the seasonal trends of soil moisture. Numerical experiments with the offline land surface scheme demonstrated that the implementation of the Mualem-van Genuchten water retention curve has the advantage for describing the processes of soil moisture transfer in comparison with Brooks-Corey parameterization. The last one is used in the basic version of the model. The Mualem-van Genuchten function provides better computational stability of the soil model in case of soils with high coarse grain content. An important contribution to the soil moisture error in the analysis and forecast is caused by specifying an inaccurate profile of the soil granulometric composition. It indicates a need to involve state-of-the-art databases for soil characteristics